Presently, transit vehicles store video data over the course of a shift and download the video data to a storage system when they return to a depot at the end of the shift. However, the bandwidth available at the depot may be limited, especially if several vehicles are downloading the video data at the same time. As a result, only a subset of the video data is often downloaded. This subset often includes video data that has been specifically identified by an operator for downloading during the shift. However, if an operator fails to identify a scene that may be interest, that scene may be lost when subsequent video data is recorded over the previously recorded video data.